Abstract: A multi-rank circuit system utilizing a shared IO channel includes a first stage of multiple selectors coupled to input multiple digital busses, and a second stage including one or more selectors coupled to receive outputs of the first stage of selectors and to individually select one of the outputs of the first stage of selectors to one or more control circuits for IO circuits of the ranks. The system switches one of the ranks to be an active rank on the shared IO channel, and operates the first stage of selectors to select one of the digital busses to the second stage of selectors in advance of switching a next active rank to the shared IO channel.
Type:
Grant
Filed:
April 27, 2022
Date of Patent:
January 23, 2024
Assignee:
NVIDIA CORP.
Inventors:
Jiwang Lee, Jaewon Lee, Hsuche Nee, Po-Chien Chiang, Wen-Hung Lo, Abhishek Dhir, Michael Ivan Halfen, Chunjen Su
Abstract: A circuit mask optimizer utilizes a Convolutional Fourier Neural Operator (CFNO) to efficiently learn layout tile dependencies, enabling stitch-less largescale mask optimization with limited intervention of legacy tools. Litho-guided self training via a trained machine learning model provides non-convex optimization, enabling iterative model and dataset refinements at a substantial performance improvement over conventional solutions.
Abstract: A transceiver configured to communicate a burst of data bits and meta-data bits for the data bits includes data channels, auxiliary data channels, and at least one error correction channel. The transceiver includes an encoder that applies 11b7s encoding to a first number of the data bits to generate first PAM-3 symbols on some or all of the communication channels, and that applies 3b2s encoding to a second number of the data bits to generate second PAM-3 symbols on at least some of the communication channels.
Abstract: To mitigate pulse shape degradation along a signal route, the signal is driven from two ends. One end of the route is loaded and the other is relatively unloaded. The loaded route and unloaded route may traverse two different metal layers on a printed circuit board. The two routes may thus be related such that the unloaded route has less RC distortion effects on the signal than does the loaded route.
Type:
Grant
Filed:
December 20, 2021
Date of Patent:
December 26, 2023
Assignee:
NVIDIA CORP.
Inventors:
Lalit Gupta, Andreas Jon Gotterba, Jesse Wang
Abstract: Layout techniques for circuits on substrates are disclosed that address the multivariate problem of minimizing routing distances for high-speed I/O pins between circuits while simultaneously providing for the rapid provision of transient power demands to the circuits. The layout techniques may also enable improved thermal management for the circuits.
Type:
Application
Filed:
September 6, 2023
Publication date:
December 21, 2023
Applicant:
NVIDIA Corp.
Inventors:
Shuo Zhang, Eric Zhu, Minto Zheng, Michael Zhai, Town Zhang, Jie Ma
Abstract: Convergence of threads executing common code sections is facilitated using instructions inserted at strategic locations in computer code sections. The inserted instructions enable the threads in a warp or other group to cooperate with a thread scheduler to promote thread convergence.
Type:
Grant
Filed:
August 11, 2022
Date of Patent:
December 19, 2023
Assignee:
NVIDIA CORP.
Inventors:
Daniel Robert Johnson, Jack Choquette, Olivier Giroux, Michael Patrick McKeown, Mark Stephenson, Sana Damani
Abstract: Neural network-based structures for action user equipment device detection, estimation of time-of-arrival, and estimation of carrier frequency offset utilized with the narrowband physical random-access channel of wireless communication systems. The structure includes a neural network to generate predictions of active user equipment devices, and a twin neural network to generate time-of-arrival predictions for signals from the user equipment devices and carrier frequency offset predictions for signals from the user equipment devices.
Type:
Application
Filed:
March 24, 2023
Publication date:
November 23, 2023
Applicant:
NVIDIA Corp.
Inventors:
Faycal Ait Aoudia, Jakob Hoydis, Sebastian Cammerer, Matthijs Jules Van keirsbilck, Alexander Keller
Abstract: Voltage level conversion circuits include PMOS pull-down devices or NMOS pull-up devices, and inverters with outputs that determine gate voltages of these devices. The inverters are powered by moving supply voltages, for example complementary supply voltages generated from a pair of cross-coupled inverters. The cross-coupled inverters may implement a data storage latch with the moving supply voltages generated from the internal data storage nodes of the latch.
Type:
Grant
Filed:
July 25, 2022
Date of Patent:
November 21, 2023
Assignee:
NVIDIA CORP.
Inventors:
Walker Joseph Turner, John Poulton, Sanquan Song
Abstract: A circuit system includes an integrated circuit package mounted on a first side of a printed circuit board and a power regulator connected to power terminals of the integrated circuit package through a cutout in the printed circuit board. The power regulator draws power from the printed circuit board by way of connections on a shelf region extending beyond an area of the cutout.
Abstract: A circuit system includes an integrated circuit package mounted on a first side of a printed circuit board and a power regulator connected to power terminals of the integrated circuit package through a cutout in the printed circuit board. The power regulator draws power from the printed circuit board by way of side pins around a periphery of the cutout.
Abstract: The differential voltage output from a first reference voltage generator of a multi-rank circuit is trained on multiple ranks of the multi-rank circuit. Multiple local reference voltage generators are trained to generate reference voltages for communication on the individual ranks, where the reference voltages output by the local reference voltage generators fall within a range of the differential voltage output.
Type:
Application
Filed:
April 27, 2022
Publication date:
November 2, 2023
Applicant:
NVIDIA Corp.
Inventors:
Jiwang Lee, Jaewon Lee, Wen-Hung Lo, Michael Ivan Halfen, Abhishek Dhir, Hsuche Nee, Po-Chien Chiang
Abstract: A multi-rank circuit system utilizing a shared IO channel includes a first stage of multiple selectors coupled to input multiple digital busses, and a second stage including one or more selectors coupled to receive outputs of the first stage of selectors and to individually select one of the outputs of the first stage of selectors to one or more control circuits for IO circuits of the ranks. The system switches one of the ranks to be an active rank on the shared IO channel, and operates the first stage of selectors to select one of the digital busses to the second stage of selectors in advance of switching a next active rank to the shared IO channel.
Type:
Application
Filed:
April 27, 2022
Publication date:
November 2, 2023
Applicant:
NVIDIA Corp.
Inventors:
Jiwang Lee, Jaewon Lee, Hsuche Nee, Po-Chien Chiang, Wen-Hung Lo, Abhishek Dhir, Michael Ivan Halfen, CHUNJEN SU
Abstract: A multi-rank circuit system includes multiple transmitters each switchably coupled to a first end of a shared input/output (IO) channel and a unified receiver coupled to a second end of the shared IO channel. The unified receiver is coupled to apply a preconfigured analog reference voltage to set a differential output of the unified receiver, and further configured to apply a variable digital code to adjust the differential output according to a particular one of the transmitters that is switched to the shared IO channel.
Type:
Application
Filed:
April 27, 2022
Publication date:
November 2, 2023
Applicant:
NVIDIA Corp.
Inventors:
Jiwang Lee, Jaewon Lee, Hsuche Nee, Po-Chien Chiang, Wen-Hung Lo, Michael Ivan Halfen, Abhishek Dhir
Abstract: A multi-rank system includes multiple circuit ranks communicating over a common data line to multiple data receivers, each corresponding to one or more of the ranks and each having a corresponding reference voltage generator and clock timing adjustment circuit, such that a rank to communicate on the shared data line is switched without reconfiguring outputs of either the reference voltage generators or the clock timing adjustment circuits.
Type:
Application
Filed:
April 27, 2022
Publication date:
November 2, 2023
Applicant:
NVIDIA Corp.
Inventors:
Wen-Hung Lo, Michael Ivan Halfen, Abhishek Dhir, Jaewon Lee
Abstract: A method includes generating a differential voltage from a first reference voltage generator; receiving the differential voltage at a second reference voltage generator; dividing the differential voltage at the second reference voltage generator into multiple available reference voltage levels; and selecting one of the available reference voltage levels to apply to a circuit.
Type:
Application
Filed:
April 27, 2022
Publication date:
November 2, 2023
Applicant:
NVIDIA Corp.
Inventors:
Jiwang Lee, Jaewon Lee, Po-Chien Chiang, Hsuche Nee, Wen-Hung Lo, Michael Ivan Halfen, Abhishek Dhir
Abstract: A machine memory includes multiple memory cells. Word lines, each with at least one word line driver, are coupled to the memory cells along rows. The word line drivers of at least some adjacent pairs of the word lines are coupled together by a pull-down transistor, in a manner that reduces read disturb of the memory cells.
Type:
Grant
Filed:
June 17, 2021
Date of Patent:
October 31, 2023
Assignee:
NVIDIA CORP.
Inventors:
Lalit Gupta, Andreas Jon Gotterba, Jesse Wang, Stefan P Sywyk
Abstract: An electrostatic discharge protection circuit is disclosed. It comprises a stacked drain-ballasted NMOS devices structure and a gate bias circuit. The gate bias circuit includes an inverter, a first gate bias output terminal, and a second gate bias output terminal. The first gate bias output terminal is coupled to a gate of a first one of the drain-ballasted NMOS devices. The second gate bias output terminal runs from an output of the inverter to a gate of a second one of the drain-ballasted NMOS devices.
Type:
Grant
Filed:
March 6, 2020
Date of Patent:
October 31, 2023
Assignee:
NVIDIA CORP.
Inventors:
Jauwen Chen, Sunitha Venkataraman, Ting Ku
Abstract: Layout techniques for chip packages on printed circuit boards are disclosed that address the multivariate problem of minimizing routing distances for high-speed I/O pins between chip packages while simultaneously providing for the rapid provision of transient power demands to the chip packages. The layout techniques may also enable improved thermal management for the chip packages.
Type:
Grant
Filed:
December 16, 2021
Date of Patent:
October 24, 2023
Assignee:
NVIDIA CORP.
Inventors:
Shuo Zhang, Eric Zhu, Minto Zheng, Michael Zhai, Town Zhang, Jie Ma
Abstract: A switch architecture enables ports to stash packets in unused buffers on other ports, exploiting excess internal bandwidth that may exist, for example, in a tiled switch. This architecture leverages unused port buffer memory to improve features such as congestion handling and error recovery.
Type:
Grant
Filed:
July 16, 2021
Date of Patent:
October 24, 2023
Assignee:
NVIDIA Corp.
Inventors:
Matthias Augustin Blumrich, Nan Jiang, Larry Robert Dennison
Abstract: Self-supervised machine learning is applied to combinational gate sizing based on an input circuit netlist. A transformer neural network architecture is disclosed to select gate sizes along paths of the network between primary inputs/outputs and/or sequential logic elements. The gate size selections may be optimized along dimensions such as path delay, path power consumption, and path circuit area.